{smcl} {* version 1.0.1 28mar2011} {cmd:help misum} {hline} {title:Title} {p 5} {cmd:misum} {hline 2} Summary statistics in MI dataset {title:Syntax} {p 8 44} {cmd:misum} [{varlist}] {ifin} {weight} [{cmd:, m(}{it:numlist}{cmd:)} {opt d:etail} {cmd:{ul:for}}{cmd:mat(}{help format:%fmt}{cmd:)} {opt mat:rix(name)}] {p 12} {helpb by} is allowed {p 12 12}{cmd:aweights}, {cmd:fweights} and {cmd:iweights} are allowed. However, {cmd:iweights} are not allowed with {opt detail}. See {help summarize} {title:Description} {pstd} {cmd:misum} calculates summary statistics in MI datasets. The program combines results from {help summarize}, applying Rubin's combination rules. {title:Remarks} {pstd}{cmd:misum} requires mi data to be {it:flong} {help mi_styles:{it:style}}. To change the style of mi data use {help mi convert} {it:flong}. {title:Options} {dlgtab:Options} {phang} {opt m(numlist)} combines results from imputed datasets {it:numlist}. {phang} {opt d:etail} calculates additional statistics. Median, skewness and kurtosis are displayed for each variable in an additional matrix. {phang} {opt for:mat(%fmt)} displays results in specified format. {phang} {opt mat:rix(name)} returns result matrix in {cmd:r(}{it:name}{cmd:)}. If {opt detail} is also specified, matrix {cmd:r(}{it:name}{hi:{it:_d}}{cmd:)} is returned additionaly. {title:Example} . sysuse auto (1978 Automobile Data) . mi set flong . mi register imputed rep78 . mi impute regress rep78 price mpg weight foreign ,add(5) [output omitted] {cmd:. misum price weight rep78} m=1/5 data Variable | Mean SD min max N -------------+------------------------------------------------------- price | 6165.257 2949.496 3291 15906 74 weight | 3019.459 777.1936 1760 4840 74 rep78 | 3.405173 .9874648 1 5.105155 74 {cmd:. misum rep78 ,detail} m=1/5 data Variable | Mean SD min max N -------------+------------------------------------------------------- rep78 | 3.405173 .9874648 1 5.105155 74 Variable | p50 Skewness Kurtosis -------------+--------------------------------- rep78 | 3 -.0418277 2.654632 . return list scalars: r(rep78_p99) = 5.105155181884766 r(rep78_p95) = 5 r(rep78_p90) = 5 r(rep78_p75) = 4 r(rep78_p25) = 3 r(rep78_p10) = 2 r(rep78_p5) = 2 r(rep78_p1) = 1 r(rep78_kurtosis) = 2.654631682655315 r(rep78_skewness) = -.0418276860909651 r(rep78_p50) = 3 r(rep78_sum) = 251.9827878952027 r(rep78_sum_w) = 74 r(rep78_Var) = .9753773811180893 r(rep78_N) = 74 r(rep78_max) = 5.105155181884766 r(rep78_min) = 1 r(rep78_sd) = .9874648279159382 r(rep78_mean) = 3.40517280939463 {title:Saved results} {pstd} {cmd:misum} calls {help summarize} and saves any results returned for each variable. It therefore saves the following in {cmd:r()}: {pstd} Scalars{p_end} {cmd:r({it:varname_stat})} {it:stat} returned by {help summarize} for {it:varname} {pstd} Matrices{p_end} {cmd:r({it:name})} result matrix ({opt matrix} only) {title:Author} {pstd}Daniel Klein, University of Bamberg, klein.daniel.81.@gmail.com {title:Also see} {psee} Online: {helpb mi}, {help summarize}{p_end}